We conducted a fragment-based screening study against IRAK4 using our integrated Medicinal/Computational Chemistry Platform. A surface plasmon resonance (SPR) “clean screen” of BIONET PROTAC Fragment Library was run to identify and remove fragments that bind non-specifically to the Biacore™ CM5 Sensor Chip. Then, “binding level screens” were carried out for the 571 remaining fragments to identify binders against the target protein. Top 12 hits from the binding level screens were validated by “affinity screens” to verify binders and estimate affinity (K_D). Machine learning (ML) was then used to design and optimize novel and diverse compounds from the SPR fragment hits which were then used as starting points for PROTAC design.